CN107429399A - Activation method for silicon substrate - Google Patents

Abstract

This invention relates to an activation composition for activating silicon substrates, comprising an aqueous solution containing a palladium ion source, a fluoride ion source, and at least two aromatic acids. The invention further relates to its use and, optionally, methods for the subsequent metallization of such treated substrates. These methods can be used in the manufacture of semiconductors and solar cells.

Description

Activation methods for silicon substrates

technical field

The present invention relates to the fabrication of solar cells and electronic devices, and in particular to methods of activation of silicon substrates prior to electroless metallization and formation of suicide interconnects for use in transistors (MOS, CMOS ), Memory Stick, MS and SD Card.

Background technique

Generally, silicide interconnects are formed by electroless deposition of a metal, such as nickel, on a suitable silicon substrate, followed by heat treatment to form the metal silicide. This heat treatment is commonly referred to as rapid thermal annealing (RTA). Typically, this heat treatment requires the substrate to be subjected to high temperatures in the range of 300°C to 750°C, whereby the metal diffuses into the silicon substrate after the metal silicide is formed. For metals to be deposited on a silicon substrate, the silicon substrate needs to be activated. This is the case especially for p-type doped polysilicon. N-type doped substrates can be plated directly using strongly basic electroless nickel plating baths. However, strongly alkaline media can damage substrates used in semiconductor manufacturing, such as solder masks. Therefore, the use of strongly alkaline plating baths is undesirable in the art. Therefore, it is a common technique to activate silicon substrates using compositions comprising palladium ions and hydrofluoric acid or other sources of fluoride ions as disclosed in GB 976,656. A possible mechanism for this type of activation is disclosed in US 4,297,393.

US 6,406,743 B1 relates to nickel silicide formation on polysilicon interconnects. The method disclosed therein teaches the use of a solution containing a palladium salt and high concentrations of hydrofluoric acid and acetic acid as an activation composition for polysilicon prior to nickel or nickel alloy deposition. Despite the high toxicity associated with the use of such concentrated hydrofluoric acid solutions, the use of this composition results in extremely coarse palladium seeds. In order to provide a uniform nickel silicide capping layer on such rough palladium seeds, thick nickel deposits need to be provided thereon, which in turn results in structures that are too large to be used in modern semiconductor technology (see Examples 1 and 2).

US 5,753,304 reports an activation solution especially for aluminum surfaces. The activation solution comprises, inter alia, palladium salts, alkali metal fluorides or hydrogen fluoride and carboxylic acids as complexing agents. The carboxylic acid is used in an amount of about 10 ml/l activation solution to about 100 ml/l activation solution.

US 2005/0161338 A1 discloses a method for activating silicon surfaces using an aqueous solution comprising a source of palladium and at least one acid. A wide variety of acids are useful in the present disclosure, such as mineral acids such as sulfuric acid, nitric acid, and hydrochloric acid; or organic sulfuric acids such as methanesulfuric acid; or aromatic sulfonic acids such as p-toluenesulfonic acid. However, the use of an aliphatic or aromatic acid results in very uneven coverage of the treated surface (see Examples 2 to 4).

WO 2014/128420 discloses the use of compositions comprising anionic or nonionic surfactants, gold ions and fluoride ions for activating semiconductor substrates. The use of surfactants improves the results and allows the formation of thinner nickel layers. According to this disclosure, compositions containing palladium and fluoride ions lead to uneven deposition and subsequent diffusion of the nickel layer on and into the substrate (pages 2, 14 and 15 and Table 1, entry 1). The use of gold ions in the manufacture of electronic devices is undesirable for a number of reasons, such as cost.

While these methods can provide methods for activating silicon substrates and subsequent formation of nickel silicides, they do not meet the needs of modern semiconductor manufacturing. The seed layer of the noble metal used is too rough and the distribution of the individual seeds on the surface of the silicon substrate is not uniform enough, so that an extremely thick nickel layer has to be deposited on the substrate. In addition, if the distribution of the precious metal is too coarse, individual particles of the precious metal are also large, resulting in an increase in cost since the price of the precious metal is usually high. Even more importantly, the metal or metal layer on the silicon substrate needs to be extremely thin and of uniform height. Therefore, it is required to be flat and smooth. It is therefore a prerequisite that the underlying palladium seed layer is extremely uniform and substantially free of macroagglomerated particles. This is especially disadvantageous if the palladium seed is as large (or even larger) than the desired metal or metal alloy layer to be formed on it. The metal or metal alloy layer formed thereon will additionally be formed on those large particles and will form an extremely rough (creating structures like valleys and hills) surface requiring a polishing step, before being formed at 5nm, 10nm, 20nm or 50nm This is especially the case in the case of metal or metal alloy layers within the range. This is incompatible with continued miniaturization and cost and environmental awareness in the semiconductor manufacturing industry.

purpose of invention

It is therefore an object of the present invention to overcome the disadvantages as mentioned above and to provide an activating composition and a method for its use which, inter alia, allow the formation of an extremely thin palladium seed layer with good surface coverage thereof on a silicon substrate.

It is therefore another object of the present invention to provide a silicon substrate uniformly covered with a palladium seed layer, which can be used for electroless metallization.

Another object of the present invention is to provide a method of metal silicide interconnect formation, especially nickel silicide interconnect formation, which complies with today's miniaturization requirements in the semiconductor industry.

Contents of the invention

These objects are solved by using the activation composition according to the invention. The activating composition according to the invention for activating silicon substrates is an aqueous solution comprising a source of palladium ions and a source of fluoride ions, characterized in that said activating composition comprises at least two (independently of each other) aromatic acids selected from: Aromatic carboxylic acids, aromatic sulfonic acids, aromatic sulfinic acids, aromatic phosphonic acids and aromatic phosphinic acids.

These objects are further solved by the inventive method for activating at least one silicon substrate, said method comprising the following steps in sequence:

(i) providing at least one silicon substrate;

(ii) activating at least a portion of the surface of the at least one silicon substrate using as an activating composition an aqueous solution comprising: a source of palladium ions, a source of fluoride ions, and at least two (independently of each other) selected from the group consisting of Acids: aromatic carboxylic acids, aromatic sulfonic acids, aromatic sulfinic acids, aromatic phosphonic acids and aromatic phosphinic acids.

Description of drawings

Figure 1 is a SEM picture of an n-type doped silicon substrate treated with a comparative aqueous activation composition consisting of hydrofluoric acid and palladium ions (corresponding to Example 1).

Figure 2 is a SEM picture of an n-type doped silicon substrate treated with a comparative aqueous activation composition consisting of hydrofluoric acid, acetic acid and palladium ions (corresponding to Example 2).

Figures 3 to 7 are SEM pictures of polycrystalline silicon substrates treated with an aqueous activation composition of the present invention comprising hydrofluoric acid, palladium ions, and one or more aromatic acids. 3 to 7 relate to embodiments 4a to 4e.

Figure 8 is an AFM image of an n-type doped polysilicon substrate treated with the activation solution of the present invention followed by an electroless nickel plating bath (referring to Example 5a).

detailed description

The activation composition for activating silicon substrates of the present invention is an aqueous solution comprising a source of palladium ions, a source of fluoride ions and at least two aromatic acids selected from the group consisting of aromatic carboxylic acids, aromatic sulfonic acids, aromatic sulfinic acids acids, aromatic phosphonic acids and aromatic phosphinic acids. In a preferred embodiment of the present invention, the at least two aromatic acids are selected from aromatic carboxylic acids, aromatic sulfonic acids and aromatic phosphonic acids.

It is understood that an aromatic carboxylic acid in the context of the present invention is an aromatic compound comprising at least one carboxylic acid moiety (—CO ₂ H). Similarly, aromatic sulfonic acids are aromatic compounds that contain at least one sulfonic acid moiety ( _-SO3H ). Aromatic sulfinic acids are aromatic compounds containing at least one sulfinic acid moiety (-SO2H), aromatic phosphonic acids are aromatic compounds containing at least one phosphonic _acid moiety ( _{-PO3H2 )} and aromatic secondary Phosphonic acids are aromatic compounds containing at least one phosphinic acid moiety (—PO ₂ H ₂ ).

The aromatic compound further comprises at least one cyclic hydrocarbon group, such as phenyl or naphthyl, wherein individual ring carbon atoms may be replaced by N, O and/or S, such as benzothiazolyl or pyridyl (such as commonly known as Nicotinic acid in 3-pyridinecarboxylic acid). Furthermore, individual hydrogen atoms bonded to aromatic compounds can be replaced in each case by functional groups such as amino, hydroxyl, nitro, alkyl, aryl, halogen groups (for example fluorine groups, chloro, bromo, iodo), carbonyl, esters derived from any of the above-mentioned acid moieties, and the like. The aromatic compound may also contain at least one hydrocarbon group consisting of two or more fused rings (such as phenanthrene or anthracene), as long as sufficient functional groups are attached to it to ensure water solubility in sufficiently high concentrations. The terms "hydroxy" and "hydroxyl" are used interchangeably herein as organic moieties. A moiety is sometimes referred to in the art as a residue or group.

As used in this specification and claims, the term "alkyl" refers to a hydrocarbon group having the general chemical formula _CmH2m+1 , _m being an integer from 1 to about 50. The alkyl moieties of the present invention may be linear and/or branched and they may be saturated and/or unsaturated. If the alkyl group is partially unsaturated, the corresponding general chemical formula must be adjusted accordingly. Preferably, m is in the range of 1 to 12, more preferably 1 to 8, even more preferably 1 to 4. C ₁ -C ₈ -Alkyl includes, for example, methyl, ethyl, n-propyl, isopropyl, n-butyl, isobutyl, tert-butyl, n-pentyl, isopentyl, sec-pentyl, tert- Pentyl, neopentyl, hexyl, heptyl and octyl. One or more hydrogen atoms of such alkyl groups may also be substituted by functional groups such as amino, hydroxyl, thiol, halogen groups (e.g. fluorine, chlorine, bromine, iodine), Carbonyl, carboxyl, carboxylate, phosphonate, etc.

Preferably, at least one of the acid moieties is directly bonded to one or more aryl groups of the aromatic compound via a carbon-carbon, carbon-sulfur or carbon-phosphorus single bond. More preferably, all acid moieties are directly bonded to one or more aryl groups.

In a preferred embodiment of the invention, the at least two aromatic acids are selected (independently of each other) from the aromatic acids according to formulas (I) and (II)

wherein R ¹ to R ¹⁴ are independently selected from hydrogen (-H), alkyl, aryl, halogen groups such as chlorine groups (-Cl), amino groups (-NH ₂ ), carboxylic acid moieties (-CO ₂ H), sulfonic acid moiety (-SO ₃ H), sulfinic acid moiety (-SO ₂ H), phosphonic acid moiety (-PO ₃ H ₂ ), phosphinic acid moiety (-PO ₂ H ₂ ), nitro (—NO ₂ ) and hydroxyl (—OH), with the proviso that at least ^one of R to R ¹⁴ is a carboxylic acid moiety (—CO ₂ H), a sulfonic acid moiety (—SO ₃ H), a sulfinic acid moiety ( -SO2H), a phosphonic acid moiety ( _{-PO3H2 )} or a phosphinic _acid moiety ( _{-PO2H2 )} , preferably a sulfonic acid moiety, a carboxylic acid moiety or a phosphonic acid moiety. Wherein independently of each other means that the at least two aromatic acids according to formula (I) or (II) can be selected from formula (I), are selected from formula (II), or one is selected from formula (I) And one is selected from formula (II).

Carboxylic acid moieties (-CO ₂ H), sulfonic acid moieties (-SO ₃ H), sulfinic acid moieties (-SO ₂ H), phosphonic acid moieties (-PO ₃ H ₂ ) and phosphinic acid moieties (-PO _{2 H2} ) are collectively referred to herein as "acid moieties".

Preferably, only one type of acid moiety is contained in a single aromatic acid. This means that, for example, if the aromatic acid comprises one or more carboxylic acid moieties, it is preferably free of other acid moieties such as sulfonic, sulfinic, phosphonic and phosphinic acid moieties.

Aromatic acids of formulas (I) to (II) are generally sufficiently soluble in water for use in the activating compositions of the present invention. If the solubility of the aromatic acid is insufficient, co-solvents and surfactants known to those skilled in the art can be used to increase the solubility of the aromatic acid in the activation composition.

In an even more preferred embodiment of the invention at least one aromatic acid of said at least two aromatic acids in said activation composition comprises a sulfonic acid moiety. In a second, even more preferred embodiment of the present invention, said activating composition comprises, among said at least two aromatic acids, at least one aromatic acid containing a carboxylic acid moiety. In a third, even more preferred embodiment of the present invention, the activation composition contains at least one aromatic acid comprising sulfonic acid moieties and at least one aromatic acid comprising carboxylic acid moieties.

In an aromatic sulfonic acid thus comprising at least one sulfonic acid moiety as according to formula (I) or (II), preferably ^at least ^one of the remaining R to R is a hydroxyl and/or amino group, more preferably Preferably, the remaining ^R1 to ^R14 are each independently selected from hydrogen, amino and hydroxyl. This acid in the activation composition appears to improve the surface coverage of the silicon substrates so treated.

In another even more preferred embodiment, said at least two aromatic acids to be used in the activating composition of the invention are selected from: benzoic acid, 1,2-phthalic acid (phthalic acid), 1,2 3-benzenedicarboxylic acid (isophthalic acid), 1,4-benzenedicarboxylic acid (terephthalic acid), 1,2,3-benzenetricarboxylic acid (benzenetricarboxylic acid), 1,2,4-benzenetricarboxylic acid Formic acid (trimellitic acid), 1,3,5-benzenetricarboxylic acid (trimesic acid), 1,2,4,5-benzenetetracarboxylic acid (pyrimellitic acid), 1,2,3,4, 5-benzenepentacarboxylic acid, 1,2,3,4,5,6-benzenehexacarboxylic acid (mellitic acid), 2-nitrobenzoic acid, 3-nitrobenzoic acid, 4-nitrobenzoic acid, 2, 5-Dinitrobenzoic acid, 2,6-dinitrobenzoic acid, 3,5-dinitrobenzoic acid, 2,4-dinitrobenzoic acid, 3,4-dinitrobenzoic acid, 2- Aminobenzoic acid, 3-aminobenzoic acid, 4-aminobenzoic acid, 3,4-diaminobenzoic acid, 3,5-diaminobenzoic acid, 2,3-aminobenzoic acid, 2,4-diaminobenzoic acid , salicylic acid, p-toluenesulfonic acid, 1-naphthoic acid, 2-naphthoic acid, 2,6-naphthalene dicarboxylic acid, 2-naphthalenesulfonic acid, 5-amino-1-naphthalenesulfonic acid, 5-amino-2- Naphthalenesulfonic acid, 7-amino-4-hydroxy-2-naphthalenesulfonic acid and phenylphosphonic acid.

The aromatic acid may be added to the activation composition in the form of the free acid, in the form of the corresponding salt (eg alkali metal or ammonium salt), in the form of its hydrate, or in any suitable combination as mentioned above. If applicable, it may also be added in the form of an anhydride (eg phthalic anhydride is capable of forming phthalic acid in aqueous medium).

The use of two or more aromatic acids in the activating composition of the present invention advantageously results in the formation of finer palladium particles (also known as seeds) on the surface of the silicon substrate so treated. The addition of two or more aromatic acids also results in an improved homogeneity of the palladium particles thus deposited on the surface of the silicon substrate. With only one aromatic acid or without any aromatic acid, the palladium particles are much coarser and less uniformly distributed on the silicon substrate surface (see comparative example).

The inventors have found that when two or more aromatic acids are used in the activation composition, the surface coverage of the silicon substrate is more uniform and the individual particles are smaller.

The concentration of the aromatic acid in the activation composition (in this context this means the total concentration of all aromatic acids used) is preferably between 0.1 mg/L and 1000 mg/L, more preferably between 1 mg/L and 750 mg/L , even more preferably in the range of 10 mg/L or 40 mg/L to 500 mg/L. Concentrations outside the stated ranges in some cases result in lesser inventive benefits, such as uniform coverage of surfaces treated with palladium seeds. Additionally, solubility issues with some aromatic acids may arise.

The activation composition of the present invention is an aqueous solution. The term "aqueous solution" means that the solvent in the solution has the predominant liquid medium being water. Other water-miscible liquids may be added, such as alcohols and other water-miscible polar organic liquids. Preferably, the activation composition comprises only water as solvent.

The activating compositions of the present invention comprise a source of palladium ions. The source of palladium ions can be any water soluble palladium salt or palladium complex. Preferably, the source of palladium ions is selected from the group consisting of: palladium chloride, palladium sulfate, palladium sulfate in the form of complexes of sulfuric acid, palladium nitrate and palladium acetate, more preferably selected from: palladium chloride, palladium sulfate, palladium nitrate and palladium acetate.

The concentration of palladium ions in the activation composition of the present invention is preferably in the range of 0.001 g/L to 1 g/L, more preferably 0.005 g/L to 0.5 g/L, even more preferably 0.05 g/L to 0.25 g/L.

The activating composition of the present invention further comprises a source of fluoride ions. The fluoride ion source can be any water soluble fluoride salt or any water soluble fluoride complex. Preferably, the source of fluoride ions is selected from hydrofluoric acid, ammonium fluoride and alkali metal fluorides such as potassium fluoride, sodium fluoride and lithium fluoride.

The concentration of fluoride ions present in the activating composition is preferably in the range of 0.075% to 4% by weight, more preferably 0.1% to 2% by weight, even more preferably 0.15% to 1% by weight. It is known in the art that higher concentrations of fluoride ions will lead to increased dissolution of silicon atoms from the substrate to be treated with the activation composition of the present invention. For the treatment of extremely thin and fragile silicon substrates, it is therefore advantageous to use lower concentrations of fluoride ions in the activating compositions of the present invention.

The activating composition of the present invention preferably has a pH of 7 or lower, more preferably lower than 3, even more preferably 0 to 2.5.

The activating composition of the present invention optionally comprises methanesulfonic acid and/or mineral acids selected from the group consisting of sulfuric acid, hydrochloric acid, nitric acid, phosphoric acid, methanesulfonic acid, hydrobromic acid, hydrogen iodide, perchloric acid, aqua regia, chlorous acid acid, iodic acid and nitrous acid. Preferably, the optional mineral acid is selected from sulfuric acid, hydrochloric acid and nitric acid. The concentration of the optional mineral acid (or methanesulfonic acid) ranges from 0.01% to 20% by weight or preferably from 0.1% to 10% by weight.

The activating composition of the present invention optionally comprises an oxidizing agent selected from molecular oxygen, a nitrate source such as potassium nitrate, and hydrogen peroxide. Molecular oxygen may be added to the activation composition in the form of a gaseous feed. It is known in the art that the optional addition of oxidizing agents, methanesulfonic acid and/or mineral acids can lead to an accelerated activation process. However, this is not always desired.

The activating composition of the present invention optionally comprises a surfactant (also known in the art as a wetting agent) selected from cationic, nonionic and anionic surfactants.

The activating composition of the invention can be prepared by dissolving all components in an aqueous liquid medium, preferably water.

The method of the present invention for activating at least one silicon substrate comprises the following steps in sequence:

(i) providing at least one silicon substrate;

(ii) activating at least a portion of the surface of the at least one silicon substrate with the activating composition of the invention.

The at least one silicon substrate to be used in the method of the invention comprises silicon, silicon oxides composed of, for example, polysilicon (including doped polysilicon, for example p-doped polysilicon and n-doped polysilicon) and monocrystalline silicon. , silicon nitride and silicon oxynitride surface(s). The entirety of the silicon substrate may be made of any of the above mentioned materials or a combination thereof or it may comprise only a surface made of one or more of the above mentioned materials.

Doping of polysilicon is typically performed using a donor (such as arsenic or phosphorus) that yields n-type doped polysilicon and an acceptor (such as boron or aluminum) that yields p-type doped polysilicon. Typically, these donors/acceptors are used at levels between 10 ⁻⁴ and 10 ⁻⁹ wt%. If very high amounts of dopants are used (typically between 10 ⁻³ and 10 ⁻⁴ wt %), n- and p-type doped polysilicon is obtained. In the context of the present invention, n- and p-doped polysilicon is also understood to mean n- and p-doped polysilicon. Silicon oxide, silicon nitride and silicon oxynitride can be doped in a similar manner as described above.

Preferably, polysilicon, p-type doped polysilicon and n-type doped polysilicon are used in the method of the invention; more preferably p-type doped polysilicon is used.

In view of activating at least a part of the surface of the at least one silicon substrate, contacting the substrate or the surface of the substrate with an activation composition according to the invention (in step (ii) ) and thus the surface of the at least one silicon substrate is activated. Contacting between the surface of the silicon substrate and the activating composition includes, inter alia, immersing the silicon substrate in the composition or wiping, spraying or otherwise introducing the activating composition onto the surface.

After the silicon substrate is contacted with the activating composition of the present invention, a thin and uniformly dispersed palladium seed layer will form on the surface of the silicon substrate. This step is known in the art as activation. Accordingly, a substrate so treated is referred to as "activated".

The substrate is activated with the activating composition for 1 second to 30 minutes, preferably 30 seconds to 10 minutes, more preferably 40 seconds to 5 minutes, most preferably 45 seconds to 2 minutes. Depending on the desired properties of the activated substrate, contact durations outside the ranges stated above may be applied.

The activation solution preferably has a temperature in the range of 10°C to 90°C, more preferably 15°C to 50°C, when in contact with the silicon substrate.

It should be understood that in the context of the present invention, a silicon substrate comprising a number of doped regions (such as source and drain, S/D), insulating layers (such as doped and undoped silicon oxide) and Semiconductor substrates with conductive layers (e.g. doped and undoped polysilicon, metals). The method is also applicable to monocrystalline or polycrystalline silicon for the manufacture of solar cells.

The method for activating a silicon substrate of the present invention may comprise another step after step (ii)

(iii) Electroless plating of metals or metal alloys on activated silicon substrates.

Electroless plating is the controlled autocatalytic deposition of a continuous metal film without the aid of an external electron supply. The main components of an electroless metal plating bath are a source of metal ions, a complexing agent, a reducing agent and as optional ingredients stabilizers, grain refiners and pH adjusters (acids, bases, buffers). Complexing agents (also known in the art as chelating agents) are used to chelate the metal to be deposited and prevent the metal from precipitating from solution (ie, as a hydroxide, etc.). Chelating a metal makes available to the metal a reducing agent that converts the metal ion to its metallic form. Another form of metal deposition is immersion plating. Immersion plating is another deposition of metal without the aid of externally supplied electrons and without chemical reducing agents. The mechanism relies on the replacement of metal from the underlying substrate by the metal ions present in the immersion solution. In the context of the present invention, electroless plating is primarily to be understood as autocatalytic deposition with the aid of a chemical reducing agent (referred to herein as "reducing agent").

To adjust the properties of the electroless metal plating bath and the metal or metal alloy deposit to be formed when such an electroless plating bath is used, additives are added to the electroless plating bath to modify the electroless plating bath and the metal or metal alloy deposit formed the nature of both. In general, electroless metal plating baths are known in the art for many metals and metal alloys.

In step (iii) a metal or metal alloy is deposited on the activated silicon substrate. Preferably the metal or metal alloy deposited in step (iii) is selected from copper, cobalt, nickel, copper alloys, cobalt alloys and nickel alloys. A more preferred embodiment of the invention is that the metal to be deposited is nickel or its alloys, since nickel suicide is a suitable replacement for tantalum or titanium suicide in chip fabrication. Electroless metal plating baths capable of depositing nickel or nickel alloys are referred to herein as electroless nickel plating baths.

The electroless nickel plating bath contains at least one source of nickel ions, which can be any water-soluble nickel salt or other water-soluble nickel compound. Preferably the source of nickel ions is selected from nickel chloride, nickel sulfate, nickel acetate, nickel methanesulfonate and nickel carbonate. The nickel ion concentration in the electroless nickel plating bath is preferably at 0.1g/l to 60g/l (0.0017mol/l to 1.022mol/l), more preferably 2g/l to 50g/l (0.034mol/l to 0.852mol/l 1), even more preferably in the range from 4 g/l to 10 g/l (0.068 mol/l to 0.170 mol/l).

The electroless nickel plating bath further contains a reducing agent selected from the group consisting of hypophosphite compounds such as sodium hypophosphite, potassium hypophosphite, and ammonium hypophosphite; boron-based reducing agents such as aminoboranes (such as dimethylaminoborane (DMAB)), alkali metal borohydrides (eg NaBH4, _{KBH4 )} ; formaldehyde; hydrazine and mixtures thereof. The reducing agent concentration (here it means the total amount of reducing agent) in the electroless nickel plating bath is usually in the range of 0.01 mol/l to 1.5 mol/l.

The pH of the electroless nickel plating bath is preferably in the range of 3.5 to 8.5, more preferably 4 to 6. Since the plating solution tends to become more acidic during its operation due to the formation of H ₃ O ⁺ ions, it can be achieved by adding bath-soluble and bath-compatible alkaline species (such as sodium, potassium or ammonium hydroxides, carbon acid and bicarbonate) for periodic or continuous pH adjustment. The stability of the operating pH of the plating solution can be improved by adding buffer compounds such as acetic acid, propionic acid, boric acid or the like in amounts up to 30 g/l, more preferably 2 g/l to 10 g/l.

In one embodiment of the invention, carboxylic acids, polyamines and sulfonic acids or mixtures thereof are selected as complexing agents. Useful carboxylic acids include mono-, di-, tri- and tetra-carboxylic acids. The carboxylic acid can be substituted with a variety of substituent moieties such as hydroxyl or amino groups, and the acid can be introduced into the electroless nickel plating bath in the form of its sodium, potassium or ammonium salt. Some complexing agents, such as acetic acid, can also be used as buffers, and the appropriate concentration of this added component can be optimized for any plating solution in view of the dual function of the component.

Examples of such carboxylic acids used as complexing agents include: iminosuccinic acid, iminodisuccinic acid, derivatives thereof and salts thereof as disclosed in WO 2013/113810; monocarboxylic acids such as acetic acid, glycolic acid , aminoacetic acid, 2-aminopropionic acid, 2-hydroxypropionic acid (lactic acid); dicarboxylic acids such as succinic acid, aminosuccinic acid, hydroxysuccinic acid, malonic acid, hydroxysuccinic acid, tartaric acid, malic acid; three carboxylic acids such as 2-hydroxy-1,2,3-propanetricarboxylic acid; and tetracarboxylic acids such as ethylenediaminetetraacetic acid (EDTA). Most preferably the complexing agent is selected from monocarboxylic and dicarboxylic acids. In one embodiment, a mixture of two or more of the complexing agents described above is utilized. The concentration of the complexing agent present in the electroless nickel plating bath or (in the case of using more than one complexing agent) the concentration of all complexing agents together is preferably between 0.01 mol/l and 2.5 mol/l, more preferably 0.05 mol/l to 1.0 mol/l range.

The electroless nickel plating bath optionally contains at least one stabilizer. Such stabilizers are required to provide adequate bath life, reasonable plating rates and control of the phosphorus or boron content in the nickel alloys so deposited. Suitable optional stabilizers are, but are not limited to, heavy metal ions such as cadmium, thallium, bismuth, lead and antimony ions; iodine-containing compounds such as iodide and iodate; sulfur-containing compounds such as thiocyanate, sulfur Urea and mercaptoalkanesulfonic acids such as 3-mercaptopropanesulfonic acid or the corresponding disulfides derived therefrom as disclosed in WO 2013/013941; and unsaturated organic acids such as maleic acid and itaconic acid; or Suitably substituted alkynes such as those taught by EP 2 671 969 A1. It is also within the scope of the present invention to use combinations of stabilizers, such as bismuth ions with mercaptobenzoic, mercaptocarboxylic and/or mercaptosulfonic acids, as taught by WO2013/113810. Electroless nickel plating baths may contain plating rate modifiers such as those disclosed in European Patent Application EP 14198380.9 to advantageously reduce the plating rate and stabilize the bath. The concentration of said at least one optional stabilizer in the electroless nickel plating bath is in the range 0.1 mg/l to 100 mg/l, preferably 0.5 mg/l to 30 mg/l.

The electroless nickel plating bath may, but need not, contain other conditioning agents such as wetting agents, surfactants, accelerators, brighteners, grain refining additives, and the like. These components are known in the art.

In case a hypophosphite compound is used as reducing agent for nickel, an alloy deposit containing nickel and phosphorus is obtained. The amount of phosphorus in the alloy deposit depends inter alia on the concentrations of hypophosphite and nickel ions and optionally stabilizers in the electroless nickel plating bath. Preferably, the amount of phosphorus in the alloy deposit is in the range of 5% to 15% by weight, the balance being nickel.

In case a reducing agent based on boron is used as reducing agent for nickel, an alloy deposit containing nickel and boron is obtained. The amount of boron in the alloy deposit depends inter alia on the concentration or pH of boron-based reducing agents and nickel ions and optional stabilizers in the electroless nickel plating bath. Preferably, the amount of boron in the alloy deposit is in the range of 1% to 10% by weight, the balance being nickel.

In case one or more of hydrazine and formaldehyde is used as reducing agent for nickel, a pure nickel deposit is obtained.

The electroless nickel plating bath may optionally contain a second source of metal ions such as molybdenum, rhenium or tungsten ions. The second metal ion may preferably be present in the form of a water-soluble salt or a compound such as MoO ₂ (OH) ₂ , ReO ₂ (OH) ₂ , WO ₂ (OH) ₂ , Na ₂ MoO ₄ , Na ₂ ReO ₄ and Na ₂ WO ₄ compound and its corresponding hydrate form.

The amount of second metal ions added to the electroless nickel plating bath is preferably in the range of 0.01 mol/l to 0.2 mol/l, more preferably 0.05 mol/l to 0.15 mol/l. The amount of the second metal ion in the electroless nickel plating bath may be sufficient to achieve a concentration of the second metal in the deposited nickel alloy of 4% to 20% by weight.

Alternatively, in step (iii), copper or copper alloy may be deposited on the activated silicon substrate in the form of metal or metal alloy. In the manufacture of solar cells, copper or its alloys are usually deposited on silicon substrates. Electroless copper or copper alloy baths contain a source of copper ions, reducing agents, complexing agents and usually stabilizers. US 7,220,296; WO 2014/154702; G.O.Mallory, J.B.Hajdu, Electroless Plating: Fundamentals And Applications, reprint, American Electroplaters and Surface Finishers Society , pp. 289-295; US 4,617,205; US 2008/0223253 and in particular European patent application EP 14198380.9 are hereby incorporated by reference in their entirety and illustrate these above-mentioned compounds for use in electroless copper or copper alloy deposition and other suitable additives (in applicable concentrations and available parameters).

Cobalt and its alloys can be used as barrier layers on silicon substrates. Such barrier layers are used in chip fabrication. It is placed, for example, between a copper wire and a silicon layer and serves to inhibit migration of copper to the silicon layer. In particular, ternary cobalt-tungsten-phosphorous and cobalt-molybdenum-phosphorous alloys can be used for this purpose. Electroless cobalt or cobalt alloy baths contain a source of copper ions, reducing agents, complexing agents and usually stabilizers. US 2005/0161338, WO 2013/135396 and European patent application EP 14198380.9 describe these above mentioned compounds and other suitable additives (in applicable concentrations and available parameters) for use in electroless cobalt or cobalt alloy deposition and Incorporated herein by reference in its entirety.

The deposited metal or metal alloy layer preferably has a thickness below 150 nm, which is more preferably in the range of 1 nm to 50 nm, which is even more preferably in the range of 2 nm to 20 nm.

At least one silicon substrate or at least a portion of its surface can be contacted with the electroless metal plating bath and the activating composition by spraying, wiping, soaking, submerging, or by other suitable means. The electroless metal plating of step (iii) in the method of the invention can be carried out in horizontal, roll-to-roll, vertical, vertical conveyor or spraying equipment. A particularly suitable plating tool which can be used to carry out the method according to the invention is disclosed in US 2012/0213914 A1.

The silicon substrate may be contacted with the electroless metal plating bath for 1 second to 30 minutes, preferably 30 seconds to 10 minutes, more preferably 40 seconds to 3 minutes, most preferably 45 seconds to 3 minutes. The contact time of the silicon substrate with the activating composition and the activated silicon substrate with the electroless metal or metal alloy plating bath affects the thickness of the metal or metal alloy layer obtained. Thus, one skilled in the art can determine the duration of contact required to achieve a particular metal or metal alloy layer thickness in both steps.

The method for activating a silicon substrate of the present invention may comprise another step after step (iii)

(iv) heat treatment of silicon substrate

And thus form a metal silicide.

The at least one silicon substrate may be heat-treated at a temperature between 300°C and 750°C in an inert gas such as nitrogen or argon for 30 seconds to 60 seconds. After heat treating a silicon substrate with a metal or metal alloy layer deposited on its surface, the metal diffuses into the silicon substrate and forms a metal silicide. This heat treatment is well established in the art and is sometimes called rapid thermal annealing (often abbreviated RTA). US 6,406,743 B1 teaches various heat treatment schemes that can be used in the context of the present invention.

Any remaining metal or metal alloy remaining on the surface of the silicon substrate after heat treatment can be removed by wet chemical etching, chemical mechanical planarization (or any other suitable means), leaving a metal silicide on the silicon substrate . Means of removing metals or metal alloys are known in the art. Nickel or its alloys can be removed by two-step etching, as disclosed in US 6,406,743 B1.

The method of the invention may include further rinsing, cleaning, etching and pretreatment steps, all of which are known in the art.

The inventive method is particularly suitable for forming interconnects made of metal silicides, such as nickel silicide interconnects. These interconnects are illustratively used in the fabrication of MOS transistors, CMOS transistors, IC substrates (VLSI). Products containing this interconnect can be memory sticks (USB sticks), MS cards, SD cards, power diodes and power transistors. Alternatively, the method of the present invention can be used to form a barrier layer, such as a cobalt alloy barrier layer, on a silicon substrate. It can also be used in the metallization of silicon substrates for the production of solar cells.

An advantage of the invention is that an extremely thin and evenly distributed palladium seed layer with a layer thickness of less than 10 nm or even less than 5 nm can be obtained. These palladium seed layers then allow an ultra-thin metal or metal alloy layer, such as a nickel or nickel alloy layer, of 50 nm or 25 nm or even 15 nm or less to be deposited thereon. Due to the extremely thin and evenly distributed palladium seed layer, the metal or metal alloy layer formed thereon is flat, planar and smooth (which can be measured by eg atomic force microscopy and XRF respectively; see Example 5). The metal or metal alloy layer thus formed can then be converted to a metal suicide with no or only reduced chemical mechanical planarization steps.

The invention will now be illustrated by reference to the following non-limiting examples.

Example

Substrates were analyzed visually by SEM (Zeiss Ultra Plus, SE2 detector, accelerating voltage 3.0 kV or 5.0 kV, data given in separate figures). Surface coverage was determined using Stream software from Olympus Corporation to quantify the measured SEM pictures.

The metal (alloy) deposition thickness was measured by XRF at 5 points on each substrate using the XRF instrument Fischerscope XDV-SDD (Helmut Fischer GmbH, Germany). Layer thicknesses can be calculated from this XRF data by assuming a layered structure of the deposit.

Scanning atomic force microscope (Digital Instruments, NanoScope, equipped with a tip radius of less than 7 nm from Nanosensors ) to measure the smoothness (or roughness) of the surface, scanning size: 5 times 2×2 μm, scanning in intermittent mode. Average roughness (SA), maximum height difference ₍ ST) and _RSAI values (relative surface area increase) were obtained from these measurements and are presented using the corresponding examples below.

Embodiment 1 (comparison):

The n-type doped polysilicon substrate was immersed in an aqueous solution containing 0.1 g/l Pd ²⁺ ions (from PdSO ₄ ) and 1 wt % HF for 120 seconds at room temperature. The surface was unevenly covered by palladium particles with an average surface coverage of (32.0±1.3)% (see Figure 1).

Embodiment 2 (comparison):

The n-type doped polysilicon substrate was immersed in an aqueous solution containing 0.1 g/l Pd ²⁺ ions (from PdSO ₄ ), 1 wt % HF and 0.5 ml/l glacial acetic acid for 120 seconds at room temperature. The surface was covered unevenly by palladium particles with an average surface coverage of (32.0±1.2)% (see Figure 2).

Example 3:

An n-type doped polysilicon substrate was immersed at room temperature in 0.1 g/l Pd ^ions (from PdSO ₄ ), 1 wt % HF in water (which further contained aromatic acids as given in Table I) Medium lasts 120 seconds. Results and treatment conditions are also given in the table.

Table I: Surface coverage obtained from each activation composition.

¹ comparative example; ² embodiments of the invention

Comparative examples a., c. and d. all show significantly improved surface coverage of the silicon surface with palladium compared to comparative examples 1 and 2 which do not contain aromatic acid. However, comparative example b. using p-toluenesulfonic acid leads to even worse surface coverage than those examples. In contrast, when more than one aromatic acid is used in inventive example e., the surface coverage is good to almost complete. It should be taken into account that the immersion time is shorter in this case than in many comparative examples and the total concentration is the same as in most comparative examples. Thus, the combination of two or more aromatic acids synergistically facilitates the process and results in improved surface coverage.

Example 4:

The polysilicon substrate was immersed in an aqueous solution each containing 0.1 g/l Pd ^ions (from PdSO ₄ ), 1 wt % HF and one or more aromatic acids at the corresponding concentrations as seen in Table II below for 60 seconds or 120 seconds.

Table II: Example 4.

¹ comparative example; ² embodiments of the invention

Compared with Comparative Examples 1 and 2, the activation solution of Example 4 both improved the coverage of the substrate surface. The individual particles as seen in the SEM pictures are mostly smaller and more uniformly dispersed over the entire surface of the substrate (compared to Comparative Examples 1 and 2), but still not satisfying today's needs. A much more uniform coverage and particle distribution on the surface is obtained only when two or more aromatic acids are used in the activation composition (entries d. and e. in Table 1). It can also be seen that the individual particles obtained from inventive examples 4d. and 4e. are smaller and that the surfaces obtained from inventive examples do not substantially form smaller particles when compared to comparative examples 4a. to 4c. large agglomerated particles.

Thus, it was found that a synergistic effect results from the use of two or more aromatic acids in the activation solution.

Embodiment 5 (the present invention):

The n-type doped polysilicon substrate was immersed in the aqueous activation composition of Example 3e for 60 seconds (Example 5a) and 120 seconds (Example 5b), respectively. So activated silicon substrates were immersed at 65°C to each having a pH of 4.3 and containing 6 g/l nickel ions (provided as nickel sulfate), dicarboxylic, tricarboxylic and hydroxycarboxylic acids as complexing agents and 0.25 g/l dimethylaminoborane as a reducing agent in an electroless nickel plating bath for 600 seconds.

The silicon substrate is uniformly covered with nickel-boron alloy. See Table III for thickness and roughness.

Table III: Layer deposit thickness and roughness.

Example Pd layer thickness [nm] Ni layer thickness [nm] S _A [nm] S _T [nm] RSAI[%] 5a 4 14 2.80 34.0 4.8 5b 8 twenty two 3.14 39.6 4.1

Very thin layers of both palladium and nickel boron are obtained by the treatment of the invention. Said layers are also extremely smooth and exhibit very little roughness deviation, which is very desirable in the manufacture of current semiconductor devices. The AFM picture of Example 5a is shown in FIG. 8 .

Claims (15)

1. An activation composition for activating a silicon substrate, wherein the activation composition is an aqueous solution comprising a source of palladium ions and a source of fluoride ions, characterized in that the activation composition comprises at least two selected from the group consisting of aromatic carboxyl aromatic acids, aromatic sulfonic acids, aromatic sulfinic acids, aromatic phosphonic acids and aromatic phosphinic acids. 2. The activating composition of claim 1, wherein the at least two aromatic acids are selected from the group consisting of aromatic carboxylic acids, aromatic sulfonic acids, and aromatic phosphonic acids. 3. The activating composition according to claim 1 or 2, wherein the at least two aromatic acids are selected from aromatic acids according to formulas (I) to (II) wherein ^R to R ^are each independently selected from hydrogen, alkyl, aryl, halogen group, amino, sulfonic acid moiety, carboxylic acid moiety, phosphonic acid moiety, nitro and hydroxyl, provided that ^R to ^R in At least one is a sulfonic acid moiety, a carboxylic acid moiety or a phosphonic acid moiety. 4. The activating composition of claim 3 with the proviso that at least ^one of R1 to ^R14 is a sulfonic acid moiety. 5. The activating composition according to any one of claims 3 or ⁴ , with the proviso that at least ^one of R to R is a carboxylic acid moiety. 6. The activation composition of claim 3, wherein one of the at least two aromatic acids is provided that ^at least ^one of R to R is a sulfonic acid moiety, and the at least two The other of the aromatic acids is provided that ^at least ^one of R to R is a carboxylic acid moiety. 7. The activating composition according to any one of claims 1 or 2, wherein the at least two aromatic acids are selected from the group consisting of benzoic acid, 1,2-phthalic acid (phthalic acid), 1,3 -phthalic acid (isophthalic acid), 1,4-benzenedicarboxylic acid (terephthalic acid), 1,2,3-benzenetricarboxylic acid (benzenetricarboxylic acid), 1,2,4-benzenetricarboxylic acid (trimellitic acid), 1,3,5-benzenetricarboxylic acid (trimesic acid), 1,2,4,5-benzenetetracarboxylic acid (pyrimellitic acid), 1,2,3,4,5 -benzenepentacarboxylic acid, 1,2,3,4,5,6-benzenehexacarboxylic acid (mellitic acid), 2-nitrobenzoic acid, 3-nitrobenzoic acid, 4-nitrobenzoic acid, 2,5 -Dinitrobenzoic acid, 2,6-dinitrobenzoic acid, 3,5-dinitrobenzoic acid, 2,4-dinitrobenzoic acid, 3,4-dinitrobenzoic acid, 2-amino Benzoic acid, 3-aminobenzoic acid, 4-aminobenzoic acid, 3,4-diaminobenzoic acid, 3,5-diaminobenzoic acid, 2,3-aminobenzoic acid, 2,4-diaminobenzoic acid, Salicylic acid, p-toluenesulfonic acid, 1-naphthoic acid, 2-naphthoic acid, 2,6-naphthalene dicarboxylic acid, 2-naphthalenesulfonic acid, 5-amino-1-naphthalenesulfonic acid, 5-amino-2-naphthalene Sulfonic acid, 7-amino-4-hydroxy-2-naphthalenesulfonic acid and phenylphosphonic acid. 8. The activating composition according to any one of the preceding claims, wherein the concentration of the at least two aromatic acids in the activating composition ranges from 0.1 mg/L to 1000 mg/L. 9. The activating composition according to any one of the preceding claims, wherein the concentration of the at least two aromatic acids in the activating composition ranges from 1 mg/L to 750 mg/L. 10. The activating composition according to any one of the preceding claims, wherein the activating composition comprises methanesulfonic acid and/or a mineral acid selected from the group consisting of sulfuric acid, hydrochloric acid, nitric acid , phosphoric acid, methanesulfonic acid, hydrobromic acid, hydrogen iodide, perchloric acid, aqua regia, chlorous acid, iodic acid and nitrous acid. 11. A method for activating at least one silicon substrate, said method comprising the steps in sequence (i) providing said at least one silicon substrate; (ii) activating at least a part of the surface of the at least one silicon substrate with an activating composition according to any one of claims 1 to 10. 12. The method according to claim 11, characterized in that the silicon substrate comprises a surface made of: silicon oxide, polysilicon, p-type doped polysilicon, n-type doped polysilicon, silicon nitride and silicon oxynitride. 13. The method according to any one of claims 11 or 12, characterized in that the method comprises a further step after step (ii) (iii) electroless plating of a metal or metal alloy on the activated said silicon substrate. 14. The method of claim 13, wherein the metal to be deposited is selected from the group consisting of copper, cobalt, nickel, copper alloys, cobalt alloys and nickel alloys. 15. The method according to any one of claims 11 to 14, characterized in that the method further comprises after step (iii) (iv) heat treatment of the silicon substrate And thus form metal silicide.